The standard way to assess a patient's response to chemotherapy is to use computed tomography (CT) to measure tumor size using uni-dimensional (RECIST) or bi-dimensional (WHO) criteria. The same standards are used to gauge the effectiveness of investigational therapies in patients with lung cancer. This methodology has changed little in the past 30 years despite the emergence of new therapies and advances in imaging technology. We and others have found that only determining the changes in the size of tumors in one or two dimensions does not adequately capture the effects of novel therapies on primary tumors and metastases. Radiographic changes in tumors treated with epidermal growth factor receptor tyrosine kinase inhibitors such as gefitinib or erlotinib or inhibitors of angiogenesis such as bevacizumab do not necessarily occur at the same magnitude or speed as observed in those individuals treated with standard cytotoxic therapies. With these newer agents, tumors respond by undergoing cystic change, central necrosis and density changes that may not be captured by conventional measurements of the largest lesion diameter. Functional changes that can only be captured on PET FDG or FIT imaging also occur. The goals of this study are to investigate the use of imaging to assess early response and progression in groups of patients receiving different types of systemic therapy (standard cytotoxic, EGFR-TKI and antiangiogenic standard) for lung cancer by correlating tumor response using advanced imaging techniques with biomarkers as well as with disease-free progression and survival. Specifically, 1.To evaluate categorical response assessment using RECIST, bi-dimensional, volumetric measures on CT and changes in metabolism on FDG-PET and proliferation on FLT-PET in patients on cytotoxic, EGFR-TKI and antiangiogenic therapy and to correlate the Best Overall Response and Progression Free Survival derived from these measurements of response with overall survival. 2. To compare total plasma DNA results with image based response (best overall response and progression free survival) and overall survival. 3. To determine if the use of a continuous scale best overall response better predicts than the categorical response assessment currently used.